Creation of Image Database with Synchronized IMU Data for the Purpose of Way Finding for Vision Impaired People

Abstract

This paper describes an image database which includes synchronized inertial measurement unit (IMU) data with the meta data of the captured images [1]. Images are taken under a range of conditions including low light, shadow conditions and controlled blurring. Physical locations are fixed and repeatable, and include accurate GPS positioning. The standardized images are synchronized over the exposure time with multiple sensor data (accelerometer, gyroscope and ambient light). This database will be used for a research project currently being undertaken at Curtin University which proposes a form of “crowd sourcing” to construct maps for use in mobility and navigation for people with vision impairments.